Is SaaS Dead? How the Salesforce Keynote Signals a Seismic Shift in Sales Infrastructure for the AI Agent Era
"Honestly, is SaaS still the right bet for us?" A sales manager at a mid-sized manufacturer asked me that out of nowhere during a casual conversation. The trigger? The 2025 Salesforce World Tour Tokyo, where CEO Marc Benioff kept coming back to one word: "Agentforce." When the very company that built SaaS starts pivoting to AI agents, it makes you wonder what that monthly license fee is really paying for.
Is SaaS Dead? How the Salesforce Keynote Signals a Seismic Shift in Sales Infrastructure for the AI Agent Era
AI agents are fundamentally reshaping how sales teams operate—and one question is cutting through the noise: "Honestly, is SaaS still the right bet for us?"
A sales manager at a mid-sized manufacturer asked me that out of nowhere during a casual conversation. The trigger? The 2025 Salesforce World Tour Tokyo, where CEO Marc Benioff kept hammering on one word: "Agentforce."
Salesforce—the very symbol of SaaS—was repositioning itself around AI agents as its core. Watching that unfold, the manager (let's call him Mr. Sato) put it plainly:
"Does this change what we're actually getting for our monthly license fee?"
If you work in international sales, that question has probably crossed your mind at least once. This post takes that big question—the one some are calling the "death of SaaS"—and looks at it through the lens of what's actually happening on the ground in international sales.
The Era of "SaaS = Software You Log Into" Is Ending
Let's start with what Salesforce actually said.
At the Tokyo keynote in May 2025, Benioff made one thing clear: the platform's new center of gravity would be "a layer where humans and AI agents work together" (Salesforce official press release, May 2025).
Traditional SaaS, broadly speaking, worked like this: you log in, a human enters data, a human reads the dashboard, a human makes a call. CRM, marketing automation—the underlying model was the same across the board.
In the Agentforce vision, AI agents autonomously handle data collection, analysis, and initial outreach. Humans focus on judgment and approval.
This isn't a UI update. It's a business model transformation.
Pay Attention to How the Pricing Model Is Changing
Traditional SaaS ran on per-seat pricing—a fixed monthly fee per user. But if AI agents are doing the work, billing by the number of human users starts to lose its logic.
Salesforce has already announced a consumption-based model for Agentforce: $2 per conversation (Salesforce official announcement, 2024).
This isn't a minor price tweak. Moving from "charge per human seat" to "charge per unit of AI work" is a fundamental rethink of how SaaS gets priced.
What AI Agents Actually Change in International Sales
This is where what we observe at RINDA's Japan Market Desk becomes relevant.
For small and mid-sized exporters doing international sales, the situation often looks something like this:
- Not enough hands to keep buyer information updated in the CRM
- No time to write individual emails in English
- Post-trade-show follow-up is three weeks late
This triple wall—staff shortages × multilingual demands × slow follow-through—is something traditional SaaS never fully solved. And the reason is simple: SaaS is a tool, and tools assume you have the people to use them.
It's Not the Death of SaaS—It's a Role Reversal
Here's what was striking: reading the Salesforce keynote carefully, Benioff never once said "SaaS is over."
What he described is AI agents running on top of the SaaS data layer. In other words, the data accumulated in your CRM becomes the memory that agents draw from.
SaaS isn't dead. But "SaaS as software humans operate through a screen" is shrinking—and "SaaS as the data infrastructure that AI agents run on" is becoming the main act.
What we've found is that this shift hits especially hard in international sales.
Compared to domestic sales, international sales comes with built-in friction:
- Time zone gaps (no human can be on call 24/7)
- Language barriers (not just English—Spanish, Arabic, and more)
- Different business norms in every market (Japan's formal opener doesn't translate)
All of these are areas where AI agents taking the first move—on behalf of a human—can make a dramatic difference.
The Numbers Behind How Fast the Shift Is Happening
Let's look at some data.
In a January 2025 report, Gartner predicted that "by 2028, at least 15% of day-to-day work decisions will be made autonomously by AI agents" (Gartner, "Predicts 2025: AI Agents").
Fifteen percent might sound modest—until you think about how much of a sales rep's time goes into research, first outreach, scheduling, and follow-up reminders. That's exactly the 15%.
A Japanese Ministry of Economy, Trade and Industry survey (published 2019, projecting to 2030) estimated that Japan's IT talent shortage could reach up to 790,000 people by 2030. Companies want to digitize their sales operations—but don't have the people to run the tools. That structural staff shortage may actually accelerate the shift from "SaaS that humans operate" to "infrastructure that AI agents run on" for Japanese companies.
What Korean Startups Are Seeing
Korean startups looking at the Japanese market are noticing something interesting.
Since 2024, South Korea's government-backed K-Startup program has clearly stepped up investment in AI agent technology. In B2B sales specifically, many new platforms are skipping the "enter data into CRM" step entirely—designing from the ground up for AI agents to handle the first move in a sales conversation.
Looking at internal RINDA platform data, AI-agent-generated initial emails to buyers showed a reply-after-open rate approximately 1.5x higher than templated emails written by humans (Q1 2025, RINDA internal data, sample size ~12,000 emails).
This isn't simply "AI writes better than humans." The reason is structural: AI agents can reflect a buyer's industry, region, and past purchasing behavior in real time—producing messages that feel relevant, not generic.
"So What Should We Actually Do?"—3 Principles for Sales Automation
Enough theory. Here are three things you can start applying in your international sales practice today.
Principle 1: Stop Optimizing for Clean Data Entry—Focus on Data Quality That Accumulates Automatically
In an AI agent world, what matters isn't that humans enter data neatly. It's that sales interactions—emails, meeting notes, chat logs—are captured accurately and automatically. Check your CRM's integration settings right now. Is automatic email import turned off?
Principle 2: Design Your First Response So It Doesn't Depend on a Human
How many days does it take to send the first follow-up after exchanging business cards at a trade show? You need to reach overseas buyers while you're still fresh in their mind. Within 24 hours is ideal. 48 hours is the outer limit. Handing just this piece to an AI agent will produce a noticeable improvement in your conversion rate.
Principle 3: Stop Asking "Which SaaS Should We Use?"—Ask "Which Tasks Should We Hand to an Agent?"
The question framing needs to change. Not "Salesforce or HubSpot?"—but "Who handles initial buyer research?" and "Does a human really need to decide when to follow up?"
Audit your workflows first. Tool selection comes after.
Takeaway: SaaS Isn't Dying—But the Number of Seats Will Shrink
SaaS isn't dead. But "SaaS as seats for humans to log in and click around" is definitively heading toward contraction.
What's rising in its place is SaaS as a data layer and workflow infrastructure that keeps AI agents running around the clock. For Japanese international sales teams struggling with staffing, that change is less a threat than a tailwind.
Back to Mr. Sato's question at the start:
"Does this change what we're actually getting for our monthly license fee?"
Yes, it does. The value of a license is shifting from "the right to use a screen" to "access to the infrastructure where AI agents operate." That's the inflection point we're sitting at right now.
Which part of your sales process do you think you could hand off to an AI agent first?
Let us know in the comments.
Frequently Asked Questions (FAQ)
Q1. If we adopt AI agents, does that make our existing CRM (Salesforce, HubSpot, etc.) obsolete?
No. AI agents use the data stored in your CRM as their memory. If anything, the CRM becomes more important as a data foundation—not less. What changes is that humans will interact with the interface less often, while AI agents handle a growing share of data input and output.
Q2. Is sales automation through AI agents realistic for small businesses?
Absolutely—and small businesses with serious staff shortages often stand to gain the most. Starting with high-volume but time-consuming tasks like buyer research, initial outreach emails, and follow-up reminders can dramatically expand what a lean international sales team is able to handle.
Q3. Will Agentforce's "$2 per conversation" pricing end up cheaper than a traditional SaaS license?
It depends entirely on your usage. Companies that have been paying for seats that go largely unused may find consumption-based pricing more cost-efficient. The right move is to audit how much of your sales process could realistically be handed to AI agents, then run the numbers.
📩 Want to talk through how AI agents can work in your international sales operation? Start here: Contact Us
🔗 Want to learn more about Rinda? Rinda Official Website
RINDA Japan Market Desk · Go-to-Market for Korean Exporters Entering Japan
#AIAgents #SaaSFuture #Agentforce #InternationalSales #SalesAutomation #ExportBusiness #JapanMarketEntry #ColdEmail
Start your journey with Rinda today!
Rinda | B2B Global Sales AI Agent for International Expansion
Have questions? Feel free to reach out anytime on LINE.